Evaluation of impact-induced damages of composite structures using memory effects of metal-coated optical fiber sensors

Abstract

In this study, a damage evaluation concept using the memory effect of metal-coated optical fiber sensors (MCOFSs) was proposed to evaluate impact-induced damages of composite structures. Two kinds of MCOFSs, i.e., aluminum-coated (Al-coated) FBG sensors and tin-coated FBG sensors, were fabricated. For Al-coated FBG sensors, a finite element analysis (FEA) was performed to investigate the capability of producing residual strain. For tin-coated FBG sensors, the capability of producing residual strain as well as the tensile failure strength was examined by experiments. It was confirmed that the induced quantity of the residual strain by MCOFSs increased with an increase of the applied strain. Then, a theoretical model for surface bonded MCOFSs on host structures was proposed. The distributions of the permanently induced residual strain in the core of the MCOFS were calculated by analytical and numerical methods when external loads were applied to the host and subsequently removed. The results were verified by laboratory experiments using the Al-coated FBG sensors bonded on carbon fiber reinforced polymer (CFRP) composite specimens. The memory effect of the MCOFS was confirmed by presenting the relationship between the residual strains of the Al-coated FBG sensors and the maximum strains experienced by the composite specimens. The effects of the material and geometric parameters of adhesive layers on the strain transfer coefficient, the residual strains, and the sensitivity coefficient of MCOFSs were investigated by a FEA, and the results were verified by laboratory experiments using the Al-coated FBG sensors. It was found that using a very soft adhesive with high non-linearity and plasticity should be avoided when` building MCOFS systems in actual situations. Moreover, small underlying adhesive thickness as well as short bonding length is favorable to achieve better performances. A response surface model among the system parameters was proposed based on the F...